😹 Fable 5 first reviews
The Neuron · Grant Harvey · 2026-07-02
Anthropic's Fable 5 model relaunched on July 1 after a government-driven export-control shutdown, now featuring a cybersecurity classifier that routes flagged requests to Opus 4.8, with early user reviews ranging from skeptical to enthusiastic.
Appears in
Extraction
Topics: frontier-ai-modelscoding-agentsai-regulationanthropic
Claims
- Fable 5 was restored after export controls were lifted, with a new cybersecurity classifier added that routes flagged requests to Opus 4.8 instead of answering them directly.
- Paid users can access Fable 5 for up to 50% of weekly usage limits through July 7, after which it becomes pay-to-play.
- Cursor benchmarks show Fable 5 leads all models on CursorBench but is the most expensive per task.
- Some early users report that concerns about Fable 5's coding request flagging were overblown, while others find minimal practical difference from Opus 4.8 for routine work.
- The newsletter recommends using Fable 5 for ambiguous, long, or judgment-heavy tasks and delegating implementation to cheaper models.
Key quotes
The smartest move this week is use Fable as much as possible where the work is ambiguous, long, or judgment-heavy, then send its conclusions and planning and design work to cheaper models.
Ethan Mollick had the opposite read: Fable changes the job from steering every step to commissioning a finished outcome.
The model is back. Now users get to find out whether the guardrails feel like seatbelts or speed bumps.